5 ms·

I am looking for beta testers for my newly launched book-recommendation website - bookclub.ai I have build an AI book-recommendation website this month & have

I am looking for beta testers for my newly launched book-recommendation website - bookclub.ai I have build an AI book-recommendation website this month & have 30k books in my database right now. Bookclub.ai has based on genres, celebrity recommendations, popular lists and series. I have applied AI to present a visual representation of "Books More like this" in the form of mind-map feature. One of the best features is "Book-Genie" which recommends books real time based on your mood. I am a sole member of this project right now and in bootstrap phase. Would love for you to checkout my website and any feedbacks/collaborations are welcomed. Happy reading! Link - bookclub.ai My Email id (for feedbacks/collaborations) - 19aditiyadav@gmail.com

jrib 4y ago
Thanks for sharing your project. Can you speak a little about what kind of AI you are using?
- aditiyadav 4y ago
  Thanks for your comments, Allow me to redirect you to https://medium.com/@19aditiyadav/how-i-built-bookclub-ai-ai-powered-book-discovery-website-8f66679f32db https://medium.com/@19aditiyadav/how-i-built-bookclub-ai-ai-...
roxgib 4y ago
I'm interested in how large you think the database might grow? One of the biggest benefits I can see in a site like this is the possibility of discovering 'hidden gems' that aren't well know and thus hard to find, but of course you'll only find what's in the database. I'm working on different product that involves processing book content, and trying to plan for the computational requirements of millions of books and it's a lot, but I'm guessing running an language model over them would be several magnitudes more demanding. Good luck, seems like a really promising use of AI!
- aditiyadav 4y ago
  I am planning to increase my catalogue to 1Mn. books as I get more users on the website. Running a language model is challenging as it took ~22 hours for training on 30k books and I expect it to take at least a week on M1 GPU. On Size, I expect 10Mn books to be taking ~16GB storage in my database. Thanks for your comments and good luck to your data pipeline.
  - roxgib 4y ago
    That's actually not too bad, I would have guessed a lot more than that. Good luck!

jrib 4y ago

Thanks for sharing your project. Can you speak a little about what kind of AI you are using?

aditiyadav 4y ago
Thanks for your comments, Allow me to redirect you to https://medium.com/@19aditiyadav/how-i-built-bookclub-ai-ai-powered-book-discovery-website-8f66679f32db https://medium.com/@19aditiyadav/how-i-built-bookclub-ai-ai-...

roxgib 4y ago

I'm interested in how large you think the database might grow? One of the biggest benefits I can see in a site like this is the possibility of discovering 'hidden gems' that aren't well know and thus hard to find, but of course you'll only find what's in the database. I'm working on different product that involves processing book content, and trying to plan for the computational requirements of millions of books and it's a lot, but I'm guessing running an language model over them would be several magnitudes more demanding. Good luck, seems like a really promising use of AI!

aditiyadav 4y ago
I am planning to increase my catalogue to 1Mn. books as I get more users on the website. Running a language model is challenging as it took ~22 hours for training on 30k books and I expect it to take at least a week on M1 GPU. On Size, I expect 10Mn books to be taking ~16GB storage in my database. Thanks for your comments and good luck to your data pipeline.
- roxgib 4y ago
  That's actually not too bad, I would have guessed a lot more than that. Good luck!